deconSTRUCT

deconSTRUCT detects and compares similar protein (sub)structures to identify and characterize structural similarities for protein structure analysis and modeling.

Key Features:

• Secondary Structure Element (SSE) Analysis: Decomposes query structures into secondary structure elements (SSEs) and reassembles them for comparison with a tunable degree of similarity in SSE direction and sequential order.
• Hierarchical Organization: Leverages hierarchical organization of protein structural information to achieve high speed and accuracy and to deliver orders-of-magnitude improvements over established database search engines without hashing substructures into static forms.
• Backbone Atom Matching: Constructs matches at the level of backbone atoms in a post-processing step to ensure precise alignment between structures.
• Reduced Representation and Global Descriptors: Uses a reduced representation of protein structures and assigns global descriptors based on geometric features for pre-filtering stages in database searches.
• Complementary Approach to Atom-by-Atom Matching: Operates as a complementary method to traditional atom-by-atom matching by focusing on main features such as SSE directions and their sequential order to discern significant similarities.
• Transformation Matrices and Matched SSE Lists: Produces transformation matrices for rigid superposition of matched structures and lists of matched SSEs as analytical outputs.

Scientific Applications:

• Structure Modeling: Aids modeling of protein complexes and conformational changes through comparison of protein (sub)structures.
• Molecular Replacement and Annotation: Supports molecular replacement and structural annotation by identifying structurally similar proteins and substructures.
• Homology Modeling: Facilitates homology modeling of protein structure and dynamics by enabling efficient comparison of substructures with varying conformational flexibility.

Methodology:

Computationally decomposes structures into SSEs, assigns global geometric descriptors for pre-filtering in database searches, reassembles and compares SSEs based on tunable direction and sequential-order similarity, and performs backbone-atom-level matching as a post-processing rigid-superposition step.